Bug#417333: Too bad it happened again

[Olivier Berger <olivier.berger@int-edu.eu>]

Le lundi 14 mai 2007 à 11:19 +0200, Olivier Berger a écrit :
> FYI, the machine reports recoverable ECC memory errors... so I'm
> suspecting problems with the memory...
> 
> But I'm not completely sure the mem is faulty though. 
> 
> There were no reports of ECC problems with the Dell diagnostic tools
> when that happened.
> 
> We're trying to solve the memory problems with the Dell support, and
> hopefully this will go away ?
> 

FYI, the memory was indeed faulty, and we changed it for a new DIMM.

Now that hardware memory errors have been fixed, we still experience
problems.

Here's a copy of the oops messages on the console :

Trace:
[<...>] inet_gso_segment
[<...>] skb_gso_segment
[<...>] dev_hard_sta...nit
[<...>] dev_queue_xmit....ae
[<...>] neigh_resolve_...
[<...>] ip_output
[<...>] reject ... [ipt_REJECT]
[<...>] ipt_ulog_target  ... [ipt_ULOG]
[<...>] reject  ...  [ipt_REJECT]      
[<...>] ipt_do_table   ... [ip_tables]
[<...>] ipt_local_out_hook ... [iptable_filt...
[<...>] nf_iterate
[<...>] dst_output
[<...>] nf_hook_slow
[<...>] dst_output
[<...>] ip_queue_xmit
[<...>] dst_output
[<...>] do_get_write_access ... [jbd]
[<...>] __ext3_get_inode_loc ... [ext3]
[<...>] tcp_transmit_skb
[<...>] __tcp_puch_pending_frames
[<...>] tcp_sendmsg
[<...>] inet_sendmsg
[<...>] do_sock_write
[<...>] sock_writev
[<...>] do_wp_page
[<...>] __handle_mm_fault
[<...>] autoremove_wake_function
[<...>] generic_file_aio_write
[<...>] do_page_fault
[<...>] sock_writev
[<...>] do_redv_writev
[<...>] vfs_writev
[<...>] sys_writev
[<...>] sysenter_past_esp
Code: ....
EIP: [<...>] tcp_tso_segment ...
 <0>Kernel panic - not syncing: Fatal exception in interrupt

Any help much appreciated.

Best regards,
-- 
Olivier BERGER <olivier.berger@int-edu.eu> (ATTENTION : new address)
Ingénieur Recherche - Dept INF
GET/INT at Evry (http://www.int-edu.eu/)
OpenPGP-Id: 1024D/6B829EEC